>===== Original Message From C Y <smustudent1@yahoo.com> =====
>Is that intended behavior? This is just my expectations so take them
>with a large grain of salt, but in the first case I would have expected
>something like:
It's not exactly the intended behavior. I'll look at it today...
><black>integrate(</black><blue>sin(x)</blue><black>+cos(</black>
>
>and in the second
>
><black>integrate(</black><blue>sin(x)</blue>+
This is close, but I include the first parenthesis in the blue coloring.
<black>integrate</black><blue>(sin(x)</blue>
>The drawback of this scheme is that the highlighter has to tell the
>difference between 2a+sin(x+y) and 2a+b*(x+y), but off the top of my
>head I think the way things work <text>() and <text><operator>() can
>always be distinguished. But it does add an extra level to things, so
>maybe it's not worth it? The advantage is that any valid self
>contained expression is always clearly indicated, but maybe that's just
>an advantage to me.
The highlighting code should (or will be soon) be able to distinguish between
the two with no problem. The algorithm I use depends on two stacks that mark
the locations of each parenthesis in the string. Its not overly complicated
but I could of missed something.
>Niiiice. Suggestion on the dialogs for plots - perhaps all the valid
>plot slots/options can be drop down menus? That way the user doesn't
>have to know before hand what's available.
Yeah, that should be possible. Of course that means I have to find out what
all the possible options are.
*dives into the documentation*
I was also thinking of possible defaults values for the fields. Any ideas?
>One other quick question - is the speed with which the highlighting is
>updated limited by the software, or is it built into the program?
I'm not sure about that actually. When text is rewritten in a different color,
it seems to take half a second or so to appear. However, when the text is
rewritten with an underline the display happens immediately. The coloring
looks better, but happens slow. Its trivial to change the hightling method.
>Very nice work so far. I like the matrix entry :-).
Thank you. :) I made those entry dialogs out of laziness. I was working with
vectors for my Calculus 3 class and wanted an easy way to enter matrices.
This actually brings up another topic regarding the socket communication and
storage of display variables. Coming soon... ;)
>P.S. - wild idea - I don't know how used/useful xmaxima's html/help
>browser is, but is there some gtkhtml widget that could be tweaked to
>duplicate its functionality? All the menu entries in xmaxima I suspect
>will be trivial to duplicate, and afaik that only leaves the help
>browser/docs differentiating gmaxima and xmaxima in usability terms.
>(Well, unless you count embedding the openmath plot, but since gnuplot
>is now the default I doubt it's an issue.) I dunno how hard it is to
>get gtk apps to work on windows (I know xchat does it with a self
>contained install, but I'm not sure how and I doubt it uses gnet) but
>it might be interesting to be able to create a Windows installable
>gmaxima distro.
Help browser:
I'm sure it will be possible to duplicate the html/help browser functionality.
Although it may be better, considering system integration and UI standards, to
convert the documentation to actual Gnome help files. Of course that would
mean that there would be two versions of the documentation, which can be a
pain in the butt. Hrmmm, something to think about.
Usability terms:
XMaxima still has a few more menu features that GMaxima lacks. Things like the
kill features and various options. Are these used at all? Should I put them
in?
Windows:
GNet should work on Windows since GNet is just an abstraction of the IO
Channel capabilities of GLib which I think work on Windows. Someone else would
have to make the Windows installer since I don't have a Windows machine around
here. It would be nice though. I'd also love to see a Mac OS X port. ;)
Thanks for the suggestions CY!
Have a good day.
--
Wes Connell
wconnell@txstate.edu